Appendix IIL 943 of this water in passing through the condenser must be multiplied by the number of Ibs. per m. to find the total heat used. (3). Heat left in condensed steam.—If all the heat calculations were made from the temperature of the hot well, there would be no need to state this result, but, as we are measuring from 32° datum, the heat left will be Ibs. of steam per m. x (temp, condensed steam — 32) Note that, if the engine is non-condensing, a hot well or tank must be supplied, though not a condenser proper and air pump. (4). Radiation and unaccounted is found by subtraction in balance sheet. (5). Heat left in jacket water. — The jackets are drained regu- larly, and the weight of water obtained. Multiply this by the degrees of temperature above 32° to find the heat lost. (6). Heat converted into work in cylinder.—This is the most important statement, and must be arrived at with great care. Indicator cards are taken every 10 minutes, and the I.H.P. calculated from the usual formula (p. 625). Then foot Ibs. per m. 772 B.T.U. per rn. The revolutions may be taken by counter or speedometer, or both. All the time the experiments are proceeding we are measuring brake H.P. by absorption (p. 576 and 875) and are thus supplied with data for mechanical efficiency (p. 770). Finally, the results may be tabulated as follows: (.) Thennal efficiency (£) Efficiency of a perfect engine within same limits (see p. 769, where r, = temp, live steam, and r2 = temp. * hot well) (c) Relative efficiency or comparison ratio (pp. 772 and 883) (d] Mechanical efficiency. The weight of dry steam per I.H.F. and per B.H.P. per hour should also be given, and indeed are usually the only results stated for commercial purposes. A graphic statement of the data, as in Fig. 890, conduces to clearness.